Networks in epigenetics

Abstract

Networks are ubiquitous. It is only recently, we started exploring what purpose they serve in real world. Biological research supposedly has, and will be, benefited most from the network science because of inherent complexity embedded in its multilayered organization. While the metabolic networks (MNs), gene-regulatory networks (GRNs) and protein-interaction networks (PINs) are being studied extensively, networks in context of epigenetics are largely ignored. Moreover, chromatin interactions networks (CINs), i.e., physical cross-talk among chromatin loci within nucleus, are rather in its infant stage. The present thesis aims to unravel and characterize networks in epigenetics using experimental and systems approach.

We studied a CIN centered on one of the Achilles heels in mammalian development, namely H19 imprinting control region (H19-ICR). Malfunctioning of this small genomic locus has pleiotropic consequences. In particular, it breaks the barrier of parthenogenesis, predisposes the animal for cancer and hyper-sensitizes the growth factor Igf2 post-receptor signaling. To explore CIN around H19-ICR (H-CIN), we devised a proximity-ligation based high resolution assay named as Circular Chromosome Conformation Capture, abbreviated as 4C. The method identifies all unknown interacting partners of a genomic locus of choice, H19-ICR in our case. Using 4C in combination with high throughput arrays, we uncovered wide-spread cis- and trans- interactions of H19-ICR in different mouse lineages. Incorporating an error in H19-ICR, that abolishes binding of CTCF protein, revealed allele-specific epigenetic regulation of H-CIN. Further analysis suggested that H19-ICR could influence the gene expression at distance. Dedicated experiments on embryonic stem cells and in-vitro derived embryoid bodies show extensive reprogramming of H-CIN, which appears to be regulated by dynamic movement of H19-ICR itself, followed up by upregulation of proximal genes during differentiation. Genes proximal to interaction sites show general traits of being developmentally regulated. Moreover, we see a significant over-representation of imprinted domains from 13 different chromosomes that further prompted us to uncover an imprinted CIN consistently in embryonic stem cells, embryoid bodies, somatic and germ-line lineages using quantitative in-situ experiments. Interestingly, CTCF binding sites within H19-ICR determines the physical proximity among imprinted loci and also transvect the epigenetic states, namely replication timing, of other imprinted loci during germ-line development. Comparative analysis on embryonic and germ-line stem cells indicates that the transvection of multiple imprinted loci by H19-ICR requires germ-line transmission. In brief, the study of H-CIN suggests that a single genomic locus could trans-regulate the epigenetic states of other genomic loci pleiotropically. Given the fact that H19-ICR is the ancient most ICR and replication asynchrony is possibly an early epigenetic feature of imprinted genes, our results might suggest that tranvection through H-CIN might represents a possible evolutionary interplay in the establishment of imprinted clusters in placental mammals.

We further explored a possible link to CTCF/Cohesins, the popular partners in orchestrating higher order chromatin structures in cis. Analysis, however, negates a general link to these and suggests multiple mechanisms of CTCF mediated trans-interactions. As this being studied, in parallel we search for the conformational feature of CTCF protein that might facilitate its binding to diverse promoters and co-factors. We predict that CTCF contains structurally disordered regions alternating to zinc fingers and towards the open terminals suggesting a possibility of multiconformation dynamics in CTCF structure that might contributes to its multiple interactions and eventually its diverse functions. This observation was an induction for the paper IV in the thesis wherein we report intrinsic structural disorder consistently in most chromatin modifiers. The physical feasibility ascribed by structural disorder might explain involvement of chromatin remodeling factors in diverse nuclear functions.

It is increasingly being realized that CINs often associate with co-regulons, which in turn may associate with PINs. Interestingly, imprinted genes, besides their CIN, also organize their gene regulatory network known as imprinted gene network (IGN). On similar lines, we attempt to analyze their PIN that might be an indirect consequence of CIN followed by IGN. Systems analyses of available human PIN data uncovered a highly central and tightly bound network module of imprinted gene-products and their interacting partners (IGPN) dedicated to imprinted gene-function. The robustness of human interactome is significantly compromised by this network and its malfunctioning makes the human interactome vulnerable to errors. We further show association of this network with several complex disorders. The study opens up novel systems perspectives in understanding imprinted gene-function in mammals.

Thus, the thesis deciphers novel networks implicated in epigenetics. In particular, we uncovered erroneous perturbations in imprinted interactomes, i.e., H-CIN and IGPN, which helps understanding their epigenetic and functional pleiotropy in mammalian imprintome.